Demand is the Hard Part: Rebuilding OTA to Last
Congress is trying again to rebuild the Office of Technology Assessment (OTA), with bipartisan legislation that is in the works, and it should. But it should also learn from past mistakes. From 1972 to 1995, OTA provided a source and process for Congress to receive technical advice and, along the way, it earned the trust of both parties. OTA ultimately closed because of a lack of demand. Too few members ever used it. A new office will last only if it builds both trust and demand right from the start.
This week, I took a deep dive into the OTA archives, which is housed at both FAS and Princeton University, to prepare for a briefing for Congressional staff on OTA’s history and future. I read through old reports and news clippings, along with other pieces that science policy scholars have written about OTA throughout the years. I read the archives from the perspective of someone who has thought a lot about the practice of “boundary spanning” between science and policy.
In 2017, I was part of a group of practitioners convened by Angela Bednarek and Carina Wyborn, held at The Pew Charitable Trusts, to describe boundary spanning from the inside, a concept with well-established scholarly roots back to the 1970s, but a less well-defined community of practice. Out of this convening, we defined it in our 2018 paper as “work to enable exchange between the production and use of knowledge to support evidence-informed decision-making in a specific context.” Sounds quite wonky, I know. But this largely invisible function provides real value.
Boundary spanners scan both the science and the policy landscape, make determinations about what research or evidence would actually be useful and when, and build the relationships that let evidence travel in both directions. Done well, our group argued, it makes research more relevant, helps decision makers absorb new evidence, and sometimes opens policy windows that might otherwise stay shut.
But the practice of boundary spanning comes with inherent tradeoffs. Advice shapes decisions only when it is salient, credible and legitimate, as science policy scholars David Cash, William Clark and colleagues at Harvard showed. In a 2003 follow-up in the Proceedings of the National Academies of Sciences (PNAS), they found that “most efforts to enhance one succeed at the expense of the others.” OTA earned credibility and legitimacy but never salience. This will be the third pillar that a new office needs in order to succeed in a durable manner.
It Took Time, But OTA Did Ultimately Build Trust
OTA’s credibility and legitimacy took most of a decade to build. Things were bumpy in the early days. By 1977, the founding OTA director had resigned, a New York Times columnist was charging that Senator Edward Kennedy was taking over the agency, the board’s Republican vice chair had quit, and appropriators had cut the budget. Board members even had their own aides working inside OTA.
The fix was to put distance between the board and the work. The board stepped back from hiring and from choosing advisory panels, and John Gibbons, who became director in 1979, made it a priority to stay out of partisan fights. The test came in 1984, when the Pentagon asked OTA to withdraw a paper on missile defense and OTA declined. By 1988, an Associated Press headline was calling OTA a “nonpartisan player.”
Trust grew from OTA’s process. A board of six senators and six representatives, split evenly by party, governed the office, and a council of outside experts advised it. OTA answered to both worlds. Panels put experts and stakeholders who disagreed in the same room, and its reports laid out options rather than recommendations. That is what our 2018 paper described as the core of boundary spanning: earning trust by “not pushing an agenda,” and offering “multiple available options and perspectives.” By 1994, 85 percent of OTA’s major reports were requested by both parties.
It Never Created Demand from Congress
OTA’s problem was not that nobody was reading its work. It literally produced best-sellers. The public bought its reports and the Government Printing Office sold 48,000 copies in 1980 alone. The problem was inside Congress. Only the board and committee leaders could request a study, so most members never asked OTA for anything. When a new majority arrived in 1995 determined to cut Congress’s own budget, few members saw OTA as theirs to defend. Peter Blair, a former OTA assistant director, has said the agency was left “without a strong constituency.” Its director, John Gibbons, put it: “If you belong to everyone, you belong to no one.”
This supply and demand imbalance is one of the most common failures I see boundary spanners grapple with. Experts are usually eager to help, and policymakers ask questions all the time. The hard part is getting them to bring those questions to an independent source, early enough to matter, and to trust the information that comes back. And AI will not solve this problem. A model can summarize almost any report in seconds, but it can’t make a member turn to a neutral body to request information before a bill is written.
Demand also has a timing problem. One of my longtime mentors, Daniel Sarewitz, and his collaborator David Guston, both science policy scholars at Arizona State University, argued in 2002 that one of the critical weaknesses of the OTA was that it started work only when a committee asked, after a technology and the interests around it had already taken shape. By the time the reports arrived, the opportunity for evidence to shape the debate had passed.
Past as Prologue
The stars might actually be aligning for OTA this time, despite at least six failed past attempts to rebuild it between 2001 and 2023. The demand that OTA never managed to build the first time, may be here, because the conditions that created OTA are back. In the 1960s, federal spending on the space program climbed faster than members’ understanding of it, and the experts testifying were often the people who would spend the money. One committee chair admitted they had to take a great deal on faith: “We are not scientists.” By 1972, Representative Charles Mosher was telling the House that Congress was “outmanned and outgunned” by the experts in the executive agencies.
Artificial Intelligence (AI), along with other complex technologies such as quantum and biotech, recreates similar demand pressure as we saw in the 1960s, with two differences that work in a new office’s favor. The first is reach. Space was the business of only a few congressional committees; AI touches almost every one, from defense to healthcare to elections. Far more members have a reason to ask OTA for help.
The second is where the expertise sits. In 1972, Congress was catching up with the executive branch. Now it is catching up mostly with industry. Nearly 90 percent of notable AI models released in 2024 came from industry, according to Stanford’s AI Index. When the Senate held its first AI Insight Forum in 2023, the headliners were Elon Musk, Mark Zuckerberg, Sam Altman and other tech CEOs, behind closed doors.
OTA Can Meet this Moment
Together, those differences create constituency OTA never had: members across nearly every committee who need answers that don’t come from the companies with the most at stake. Whether that need can meet the moment remains to be seen and the devil will be in the details. Rebuilding OTA means building a boundary spanning function into Congress itself, as a full-time job rather than an ad hoc endeavor.
Congress can write the preconditions for trust into law: shared governance, an open process, options rather than advocacy. It can design a statute to increase the likelihood for demand: every member able to ask, answers returned fast enough to use, and a standing duty to look ahead, scanning the horizon for technology that is not yet mainstream. All of the ingredients are there. If we set up the new OTA to do its job well, members will see it as the first stop to see what issues are lurking around the corner. They will bring their questions early. And they will keep coming back for more.
If Congress needs help understanding the risks and rewards of a piece of technology, they shouldn’t have to rely solely on the sparknotes from a salesman. The rub is much simpler: Congress needs nerds.
If we set up the new OTA to do its job well, members will see it as the first stop to see what issues are lurking around the corner. They will bring their questions early. And they will keep coming back for more.
Even the best-designed policy rests on assumptions about who will staff it, how it will be funded, whether the responsible institutions have ever coordinated, and if the goals are achievable.
Congress should establish a Modeling, Analysis, and Technical Talent for Energy Regulation Hub (MATTER Hub) within the Department of Energy (DOE).